Pilot Projects

Trinity College Dublin
Céline De Looze, Ph.D., Principal Investigator
Christine McGarrigle, Ph.D.

The Harmonized Cognitive Assessment Protocol (HCAP) initiative aims to establish research diagnoses of probable presence or absence of cognitive impairment and dementia, using algorithms built on harmonized measures of cognitive function from population-based aging cohorts of the HRS-family studies, including the University of Michigan Health and Retirement Study (HRS), The Irish Longitudinal Study of Ageing (TILDA) and the English Longitudinal Study of Aging (ELSA). In this pilot project, we propose to (i) examine whether and by how much changes in activity measures collected as part of the HCAP informant interviews can help improve risk stratification of cognitive and dementia categories based on HCAP dementia research diagnosis tool and (ii) to investigate by how much they can improve proxy measures for imputation of cognitive function when data is missing.

University of Pennsylvania
Irma Elo, Ph.D., Principal Investigator

This project aims to validate the Harmonized Cognition Assessment Protocol (HCAP) in Ghana, a sub-Saharan African (SSA) low-middle-income (LMIC) country located in western Africa with the goal of integrating the HCAP into the nationally representative longitudinal SAGE survey. This project will build on a sample of respondents interviewed in the INDEPTH WHO-SAGE Adult Health and Ageing Survey (INDEPTH SAGE) implemented by the Navrongo Health Research Center (NHRC) in the catchment area of the Navrongo Demographic Health and Surveillance System (NDHSS) in January–April 2007. The project will utilize the infrastructure of the NHRC, a leading research institution established in 1988, conducting health research in Ghana. The projects will focus on the development, validation, and pre-testing of the Ghana-HCAP, an HCAP-comparable cognitive assessment protocol for SSA LMIC. In addition, based on the collected cognition data and within the framework of a diagnostic consensus conference consisting of U.S. and Ghanaian experts, the project will develop an HRS-like approach to classify respondents into three distinct groups: normal cognitive aging, cognitively impaired no dementia (CIND) and ADRD. We will conduct preliminary analyses and assess the properties of the Ghana-HCAP instrument and, through linkage to INDEPTH-SAGE from 2007, analyze the predictors of cognitive changes over time.

University of Michigan
Joshua R. Ehrlich, M.D., M.P.H., Principal Investigator

This project builds on the Longitudinal Study of Health and Ageing in Kenya (LOSHAK) to investigate the health and economic well-being of the rapidly aging population in Kenya. Currently, there is little population-level data from sub-Saharan Africa to support research on critical domains in aging, including risk factors for Alzheimer’s disease and related dementias (ADRD) and mental health, air pollution and climate vulnerability, and economic well-being. To address this gap, LOSHAK joins a growing network of population-representative harmonized panel studies of aging in 45 countries modeled on the U.S. Health and Retirement Study (HRS). This project undertakes a comprehensive and rigorous measurement of cognitive health in LOSHAK to build the capacity to field a nationally representative sample in Wave 1.

Columbia University & The University of Texas Medical Branch
Yuan Zhang, Ph.D., Principal Investigator
Phillip Cantu, Ph.D., Principal Investigator

This project examines informant reports of cognitive decline across developed and developing countries using data from HCAP studies fielded in the U.S., Mexico, and China. We will evaluate the comparability of informant assessments using data from the HCAP of the HRS and China Health and Retirement Longitudinal Study (CHARLS), and the Ancillary Study of Cognitive Aging in Mexico (Mex-Cog). We will (1) describe the study design, interview protocol, and the informant protocol in each survey and (2) identify comparable and non-comparable items in the informant interviews across the three countries. We will also compare the informant responses to comparable questions in each survey by modeling responses to each comparable question for each country, controlling for age, sex, and measured cognition. We will examine the extent to which informant characteristics and the informant-respondent relationship are associated with informant reports within and across countries.

Trinity College Dublin
Céline De Looze, Ph.D., Principal Investigator
Christine McGarrigle, Ph.D.

This pilot proposal will develop a weighting methodology that reduces bias in estimates of the incidence and prevalence of cognitive decline associated with selective attrition in The Irish Longitudinal Study on Ageing (TILDA). This methodology will be applicable to other cohort studies within the HCAP research network.

University of Pennsylvania
Irma Elo, Ph.D., Principal Investigator
Rebeca Wong, Ph.D.

This pilot study used data from the 2016 Mex-Cog (N=1,674) and the 2019 Chile-Cog (N=1,946) studies to compare socioeconomic covariates of cognition in older adults in Latin America. On average, Mex-Cog respondents showed lower scores for adapted MMSE instruments than Chile-Cog, with Mex-Cog respondents more likely to have less education and be from poorer childhood socioeconomic backgrounds than Chilean counterparts. Regression analyses revealed that gender, nationality, and schooling were also predictive of adapted MMSE scores in both Mex-Cog and Chile-Cog respondents. Current findings from the pilot study ultimately suggest more rapid age-related performance declines in Mex-Cog respondents, who express more negative effects for unobserved factors. There are intergroup similarities, such as the higher impact of schooling for women than men, in predicting total cognition scores in both countries. Future work will incorporate imputed data for Chile-Cog, explore the underlying mechanisms driving the observed effects of schooling and gender alongside comparative analysis of other explanatory variables (especially social safety nets), and investigate alternative cognitive function measures.

Trinity College Dublin
Joanne Feeney, Ph.D., Principal Investigator

Current HCAP diagnostic algorithms rely on informant reports to subsequently classify the participant as having normal cognition, mild cognitive impairment, or dementia when objective cognitive test data is unavailable. This assumes that the informant’s assessment is accurate and not subject to bias; however, this remains largely untested. We use data from the core studies and HCAP sub-studies of three longitudinal studies within the Health and Retirement Study International Network of Studies (HRS-INS) to examine if the cognitive function of spouse/partner informants predicts Informant Questionnaire of Cognitive Decline in the Elderly (IQCODE) and Blessed Dementia Rating Scale scores (BDRS) and determine whether adjusting for informant cognition improves the ability of informant scores to predict the objective cognitive function of respondents. We also investigate whether the quality of the respondent-informant relationship influences informant ratings and determine whether adjusting for relationship quality improves the performance of the IQCODE and BDRS, as above.

University College London
Shabina Hayat, Ph.D., Principal Investigator
Sarah Assaad, Ph.D.

This projects examines the design and methodology used across four English-speaking HCAP studies: ELSA, HRS, TILDA, and NICOLA. We will conduct an in-depth review of these studies to examine the extent to which the original protocol (used in HRS) has been either directly adopted or adapted in UK-based studies and the potential impact of differences in methods on the quality and comparability of the data collected. Although cross-cohort measurement differences have been examined in linguistically and culturally diverse settings, a direct comparison in English-speaking studies has not been examined. This study contributes to the aims of the HCAP Network to increase rigor and reproducibility of HCAP, strengthen evidence for causality, examine test-retest reliability and methodology within different HCAP studies, determine bounds of generalizability, and identify sources of heterogeneity and cohort effects.

The American University in Cairo
Mohamed Salama, M.B.B.C.H, D.T.Q.M., Ph.D., Principal Investigator

This projects seeks to validate the Harmonized Cognitive Assessment Protocol (HCAP) diagnostic assessments in Arabic for a subsequent Egyptian national population-based prevalence survey. The aim of AL-SEHA HCAP is to identify which interactions of biomedical and socio-economic conditions over the life course affect cognition in later life.

Washington University in St. Louis
Jorge Llibre-Guerra, M.D., M.S., Principal Investigator

This project explores dementia prevalence and risk factors in Latino populations. We expect that the relative contributions of risk factors (e.g., genetic, SDH, and comorbidities) differ within Latino subgroups which explains a differential effect in age at onset, dementia prevalence, and disease progression. We will implement an effective method for cross-national harmonization of Latin American population studies, including 10/66 inception studies and 10/66 family of studies, Mexican Health and Aging Study/ Cognitive Aging Ancillary Study in Mexico (MHAS/MexCog), and the Brazilian Longitudinal Study of Aging (ELSI). We will establish a common cognitive impairment and dementia algorithm across 10/66 studies, MHAS/MEXCOG, and ELSI. We also compare risk factors and prevalence of dementia by age, education, and sex for each contributing study.

University of Michigan
Emily Briceño, Ph.D., Principal Investigator
Miguel Arce Rentería, Ph.D., Principal Investigator

This project harmonizes cross-national HCAP data across the U.S. (HRS HCAP), Mexico (Mex-Cog), and Chile (Chile-Cog) to produce precise, clear, and comprehensive pre-statistical harmonization data of cognition and informant ratings that will lay the foundation for future researchers to perform harmonization utilizing these cohorts. We will generate and validate harmonized scores of global cognitive performance, memory, and language, in addition to informant ratings of cognitive decline. These scores will be available for future substantive investigations of the social, economic, and cultural determinants of cognitive health and cross-national prevalence estimates of mild cognitive impairment and dementia.